ABSTRACT Pancreatic ductal adenocarcinoma (PDAC) has a median 5-year survival of only 8%, and early diagnosis of PDAC is an area of highest priority for the NCI. Amongst the best-recognized risk factors for PDAC are mucinous pancreatic cysts, of which the most common subtype is known as intraductal papillary mucinous neoplasm (IPMN). Currently, IPMN patients either undergo surgical resection due to ?worrisome? imaging features, or are followed conservatively by serial imaging studies for risk of progression to invasive PDAC. Unfortunately, the imaging criteria reflexing patients to surgery are imperfect, leading to both over- and under-treatment of IPMNs. Further, there are no credentialed blood-based biomarkers with a sensitivity and specificity that warrants reliable therapeutic stratification. Our group has identified oncogenic mutations of KRAS and GNAS as the two most common driver mutations in IPMNs ? one or other is present in ~96% of cases. We have now engineered the first animal model of IPMN that harbors the mutational combination (Kras;Gnas) found most commonly in the cognate human disease. Upon doxycycline induction, the Kras;Gnas mice uniformly develop cystic lesions by 6 weeks, with progression to invasive cancer in 25% of mice by 21 weeks, mimicking the multistep progression of human IPMN to PDAC. The objective of this proposal is to enhance the translational applicability of this model by using it as a controlled platform to address key unmet needs in the management of IPMNs in two areas: imaging correlates and circulating biomarkers. In Aim 1, we will use the animal model to investigate two novel imaging platforms ? quantitative feature extraction from MRI scans using an indigenously developed algorithm known as ?Enhancement Pattern Mapping? (EPM) and second, hyperpolarized MRI (HPMRI), in order to determine imaging correlates that coincide with the transition from low grade IPMN to cancer. In Aim 2, we will use a combination of unbiased mass spectrometry and array-based approaches to identify circulating proteins and autoantibodies, respectively that correlate with progression of murine IPMNs to PDAC. In addition, we will examine the potential of genomic liquid biopsies for cancer prediction, through utilizing an ultrasensitive and quantitative droplet digital PCR (ddPCR) platform for detection of mutant KRAS and GNAS DNA within circulating exosomes. All three classes of blood-based biomarkers (proteins, autoantibodies and exoDNA) will be assessed in matched murine plasma samples, which will allow us to estimate the additive performance for cancer detection using robust statistical paradigms. Both aims will benefit from ready access to imaging scans and biospecimens from IPMN patients for cross-species translational validation studies, through NCI-funded multicenter U01 consortia that are led by the PI. We believe this multidisciplinary proposal has the potential for long-term impact on PDAC mortality through practice changing alterations in the approach towards monitoring cancer progression in IPMNs.